Transforming Government with Big Data

DEMYSTIFYING
BIG DATA

A Practical Guide To Transforming The Business of Government

Prepared by TechAmerica Foundation’s Federal Big Data Commission 1

TechAmerica Foundation: Federal Big Data Commission

Listing of Leadership and Commissioners
Leadership
Steve Mills (Co-Chair)

Steve Lucas (Co-Chair)

Senior Vice President and Group Executive
IBM

Global Executive Vice President and General Manager,
Database and Technology
SAP

Leo Irakliotis (Academic Co-Chair)

Michael Rappa (Academic Co-Chair)

Western Governors University

North Carolina State University

Teresa Carlson (Vice Chair)

Bill Perlowitz (Vice Chair)

Vice President Global Public Sector
Amazon Web Services

Chief Technology Officer, Science, Technology and Engineering
Group
Wyle

Commissioners
Manish Agarwal
Attain

Eric Gillespie
Poplicus, Inc.

Ray Muslimani
GCE

Sevag Ajemian
Globanet

John Igoe
Dell

Tim Bradley
MicroStrategy

Prem Jadhwani
GTSI Corp.

James Norton
General Dynamics
C4 Systems

Rich Campbell
EMC Corporation

Richard Johnson
Lockheed Martin IS&GS

Stephen Coggeshall
ID Analytics

Yogesh Khanna
CSC

Bill Cull
Splunk

Bilhar Mann
CA Technologies

David Shacochis
Savvis, A CenturyLink
Company
Hemant Sharma
CGI

Mike Olson
Cloudera

Rick Sullivan
Hewlett-Packard

Steven Perkins
Grant Thornton LLP

Michael Van Chau
MEI Technologies

Raghu Ramakrishnan
Microsoft

Dale Wickizer
NetApp

Rich Rosenthal
TASC

Federal Big Data
Commission Staff
Chris Wilson
Staff Director

2

TechAmerica
Foundation Staff
Jennifer Kerber
President

Demystifying Big Data: A Practical Guide To Transforming The Business of Government

Table of Contents
Listing of Leadership and Commissioners

2

Foreword

5

Executive Summary Key Findings
6

Defining Big Data7

Leveraging Big Data –

Early Experiences and Lessons Learned7

Recommendations For Getting Started –

For Agency Leaders8

Beyond the Individual Agency –

Policy Considerations8
Defining Big Data Business/Mission Value
9

Current Trends and Explosion of Big Data9

Big Data Definition10

Characteristics of Big Data10

Mission / Value of Big Data12

The Use of Big Data:

Potential Business Mission Value 13

Healthcare Quality and Efficiency13

Healthcare Early Detection13
Transportation14
Education14

Fraud Detection –

Healthcare Benefits Services14

Cyber Security15

Fraud Detection – Tax Collection15
Weather15

New Ways of Combining Information –

Helping the Unemployed Find Work15
Conclusions15
Big Data Case Studies
16

National Archive and Records

Administration (NARA)17

Royal Institute of Technology of Sweden (KTH)18

Vestas Wind Energy19

University of Ontario Institute of Technology20

NASA Johnson Space Center 21

Technology Underpinnings
22
Introduction22

Big Data Infrastructure22

Big Data Core Technologies23

Big Data Accelerators24

Integration Layer24

Networking Considerations25

Cloud Computing Considerations26

Understand Source Data

and Applications27

Data Preparation –

Cleansing Verification 27

Data Transformation27

Business Intelligence/Decision Support27

Analysts/Visualization27
The Path Forward: Getting Started
28

Observations Lessons Learned28

Recommended Roadmap

for Getting Started29
Public Policy
31

Accelerating Uptake of Big Data

in the Federal Government31

Education and Workforce Development 33

Leveraging New Talent 33

Increasing Talent33

Research and Development

Considerations34

Privacy Issues35

Removing Barriers to Use through

Procurement Efficiencies 36
Conclusion

37

Acknowledgements

38

Deputy Commissioners

39


3

Table of Illustrations
Table 1: Characteristics of Big Data 11
Table 2: Case Studies High Level Summary

16

Figure 1: Big Data Enterprise Model 22
Figure 2: Notional Information Flow – The Information Supply Chain

26

Figure 3: Road Map for Getting Started

29

Table 3: Practical Road Map Summary 30

4


Foreword

In recent years, federal, state and local government
agencies have struggled to navigate the tidal wave of
sheer volume, variety, and velocity of data that is created
within their own enterprise and across the government
ecosystem. As this tidal wave has swept across
government, “Big Data” has arisen as the new ubiquitous
term. Everyone is talking about Big Data, and how it will
transform government, both in Washington and beyond
the Beltway. Looking past the excitement, however,
questions abound. What is Big Data? What capabilities
are required to keep up? How do you use Big Data to
make intelligent decisions? How will agencies effectively
govern and secure huge volumes of information,
while protecting privacy and civil liberties? Perhaps
most importantly, what value will it really deliver to the
government and the citizenry it serves?

Although there clearly is intense focus on Big Data,
there remains a great deal of confusion regarding
what the term really means, and more importantly, the
value it will provide to government agencies seeking
to optimize service outcomes and innovate. This
confusion may be due in part to the conversation
being driven largely by the information technology
community versus line of business community, and
therefore centering primarily on technology. This
report approaches Big Data from the perspective of the
key mission imperatives government agencies must
address, the challenges and the opportunities posed
by the explosion in data, and the business and inherent
value Big Data can provide. The report breaks the
discussion down into five chapters:
1. Big Data Definition Business/Mission Value

In order to answer these questions and to provide
guidance to our federal government’s senior policy
and decision makers, the TechAmerica Foundation Big
Data Commission relied upon its diverse expertise and
perspectives, input from government representatives,
and previous reports. The Commission’s mandate
was to demystify the term “Big Data” by defining its
characteristics, describe the key business outcomes it
will serve, and provide a framework for policy discussion.

2. Big Data Case Studies
3. Technical Underpinnings
4. The Path Forward: Getting Started
5. Public Policy
The Commission based its findings on the practical
experiences of those government leaders who have
established early successes in leveraging Big Data,
and the academics and industry leaders who have
supported them. The intent is to ground report
recommendations in these best practices and lessons
learned, in an effort to cut through the hype, shorten
the adoption curve, and provide a pragmatic road map
for adoption.


5


Big Data has the potential to transform government and
society itself. Hidden in the immense volume, variety and
velocity of data that is produced today is new information,
facts, relationships, indicators and pointers, that either
could not be practically discovered in the past, or
simply did not exist before. This new information,
effectively captured, managed, and analyzed, has the power
to enhance profoundly the effectiveness of government.
Imagine a world with an expanding population but a reduced
strain on services and infrastructure; dramatically improved
healthcare outcomes with greater efficiency and less
investment; intensified threats to public safety and national
borders, but greater levels of security; more frequent and
intense weather events, but greater accuracy in prediction
and management. Imagine a world with more cars, but less
congestion; more insurance claims but less fraud; fewer
natural resources, but more abundant and less expensive
energy. The impact of Big Data has the potential to be as
profound as the development of the Internet itself.
Harnessing Big Data also will serve the key objectives and
recommendations described in the Digital Government
Strategy report the White House released on 23 May 2012 –
“Digital Government: Build a 21st Century Platform to Better
Serve The American People” (Digital Government Strategy).
A primary component of the Digital Government Strategy is
to “unlock the power of government data to spur innovation
across our nation and improve the quality of services for
the American people.” Big Data promises to fulfill the very
essence of this objective.
The great paradox is that, as Big Data emerges as a new
resource, we struggle to keep pace. We find it difficult
to discover, understand, and leverage the information it
contains, to find those true nuggets of knowledge that can
improve the lives of everyday citizens and change the
world. Although there is more data available, our ability
to comprehend this data is reduced. The challenge lies in
capturing the streams of Big Data that we need, effectively
managing them, and extracting new and relevant insights.
The good news is that not only is it possible to extract value
from Big Data, but the path is relatively straightforward.
Leaders across government, academia, and private industry
have made investments, have demonstrated success,
and we now know what “good” looks like; there exist best
practices from which we can define the path forward.
6

These experiences reveal that although the
impact of Big Data will be transformational,
the path to effectively harnessing it does not
require government agencies to start from
scratch with greenfield investment. Rather
government can build iteratively on the
capabilities and technologies it already has
in place.
Perhaps as important, the path to harnessing the value
of Big Data is now affordable. It is this convergence
of the availability of Big Data, the ability to harness it,
and the affordability of doing so that brings government
to an inflection point. The time to act is now.
Success in capturing the transformation lies in
leveraging the skills and experiences of our business
and mission leaders, rather than creating a universal
Big Data architecture. It lies in understanding a
specific agency’s critical business imperatives and
requirements, developing the right questions to ask,
understanding the art of the possible, and taking initial
steps focused on serving a set of clearly defined use
cases. The experiences and value gained in these
initial steps lead to more questions, more value, and
an evolutionary expansion of Big Data capability that
continually leverages prior investments.
It is instructive to remember the phenomenon of
eBusiness. In the late 1990s, the buzz was that
eBusiness was going to change the world. By 2005,
the term largely had faded away, it became passé.
Yet, looking across government and society, it is clear
that effective organizations operate on the fundamental
principles wholly consistent with the term. eBusiness
did in fact change the world. One can argue that those
organizations that successfully harnessed the power of
eBusiness started with their operational challenges and
requirements first, and asked, “How can the Internet
help?,” versus diving immediately into the technology.
So it will be with Big Data. Successful government
agencies will seek to define their requirements and
use cases, and ask, “How can Big Data help?” versus
setting out to deploy Big Data projects. Ten years from
now, we may have forgotten the term, but its principles
will underpin society.



Defining Big Data
• Big Data is a phenomenon defined by the rapid acceleration in the expanding volume
of high velocity, complex, and diverse types of data. Big Data is often defined along
three dimensions -- volume, velocity, and variety.
• This phenomenon represents both a challenge in making sense of the data
available to governments, and an opportunity for government agencies that seek to
exploit it to enhance the business of government.
• Addressing the challenge and capturing the opportunity requires advanced
techniques and technologies to enable the capture, storage, distribution, management,
and analysis of the information.
• Government leaders should strive to understand the “Art of the Possible” enabled by
advances in techniques and technologies to manage and exploit Big Data. Example of
use cases and live case studies are critical in understanding the potential of Big Data.

Leveraging Big Data – Early Experiences and Lessons Learned
• While Big Data is transformative, the journey towards becoming Big Data “capable” will be iterative and cyclical,
versus revolutionary.
• Successful Big Data initiatives seem to start not with a discussion about technology, but rather with a burning
business or mission requirement that government leaders are unable to address with traditional approaches.
• Successful Big Data initiatives commonly start with a specific and narrowly defined business or mission
requirement, versus a plan to deploy a new and universal technical platform to support perceived future
requirements. This implies not a “build it and they will come” transformative undertaking, but rather a “fit for
purpose” approach.
• Successful initiatives seek to address the initial set of use cases by augmenting current IT investments, but do so
with an eye to leveraging these investments for inevitable expansion to support far wider use cases in subsequent
phases of deployment.
• Once an initial set of business requirements have been identified and defined, the leaders of successful initiatives
then assess the technical requirements, identify gaps in their current capabilities, and then plan the investments to
close those gaps.
• Successful initiatives tend to follow three “Patterns of Deployment” underpinned by the selection of one Big Data
“entry point” that corresponds to one of the key characteristics of Big Data – volume, variety and velocity.
• After completing their initial deployments, government leaders typically expand to adjacent use cases, building
out a more robust and unified set of core technical capabilities. These capabilities include the ability to analyze
streaming data in real time, the use of Hadoop or Hadoop-like technologies to tap huge, distributed data sources,
and the adoption of advanced data warehousing and data mining software.


7


Recommendations For Getting
Started – For Agency Leaders
Big Data is a phenomenon characterized by the exponential
expansion of raw data that is inundating government and
society. It is already here and it is accelerating. The path
to effective government described in the Digital Government
Strategy lies in developing a set of capabilities to meet the
challenge and harness its value. Perhaps unlike any other
technical challenge the government now faces, Big Data will
not be ignored.

Beyond
the Individual Agency –
Policy Considerations
From a policy perspective, the federal government
should examine existing organizational and
technical structures to find and remove barriers
to greater Big Data uptake and, where needed,
take action to accelerate its use. Specifically, the

The question lies in how to respond. The good news is that

government should:

leveraging Big Data is affordable and early experiences
offer best practices. As many government agency leaders

1. Expand the talent pool by creating a

take the first steps toward adopting Big Data solutions, the

formal career track for line of business and

Commission makes the following recommendations:

IT managers and establish a leadership
academy to provide Big Data and related

1. Understand the “Art of the Possible” -- Explore the

training and certification.

case studies contained in this report, posted on the
TechAmerica Foundation Website, and otherwise in the

2. Leverage the data science talent by

public domain to find inspiration and practical examples.

establishing and expanding “college-togovernment service” internship programs

2. Identify 2-4 key business or mission requirements that

focused specifically on analytics and the

Big Data can address for your agency, and define and

use of Big Data.

develop underpinning use cases that would create value
for both the agency and the public.

3. Establish a broader and more long-lasting
coalition between industry, academic centers,

3. Take inventory of your “data assets.” Explore the data

and professional societies to articulate and

available both within the agency enterprise and across the

maintain professional and competency

government ecosystem within the context of the business

standards for the field of Big Data.

requirements and the use cases.
4. Expand the Office of Science and
4. Assess your current capabilities and architecture

and development strategy for Big Data

the deployment entry point that best fits your Big Data

to encourage further research into new

challenge – volume, variety or velocity. (The entry points

techniques and tools, and explore the

are described in detail in the “The Path Forward: Getting

application of those tools to important

Started” chapter of this report.)

problems across varied research domains.

5. Explore which data assets can be made open and

5. Provide further guidance and greater

available to the public to help spur innovation outside the

collaboration with industry and stakeholders

agency. Consider leveraging programs like the Innovation

on applying the privacy and data protection

Corps offered by the National Science Foundation, or the

practices already in place to current

Start-Up America White House initiative.

8

Technology Policy (OSTP) national research

against what is required to support your goals, and select

technology and cultural realities.


Defining Big Data

Defining Big Data Business/Mission Value
Big Data is not a technology, but rather a phenomenon resulting from the vast amount of raw information generated across
society, and collected by commercial and government organizations. This phenomenon represents both a challenge in
harnessing this volume of data, and an opportunity for government agencies who seek to enhance their effectiveness. This
section describes the accelerating explosion of Big Data, the definition of Big Data, the characteristics of Big Data, the
mission and business value Big Data promises, and potential use cases.

Current Trends and Explosion of Big Data
In recent years, federal, state, and local governments have come to face a tidal wave of
change as a result of the drastic increase in the sheer volume, variety and velocity of data
within their own enterprise and across the government ecosystem. For example, in 2011,
1.8 zetabytes of information were created globally, and that amount is expected to double
every year. This volume of data is the equivalent of 200 billion, 2-hour HD movies, which
one person could watch for 47 million years straight. The impact of this phenomenon to
business and government is immediate and inescapable.
Because of the Internet and influx of information from multiple sources embedded within
every fabric of our government, agencies will continue to struggle with managing large
streams of data. Our government has access to a constant barrage of data from sensors,
satellites, social media, mobile communications, email, RFID, and enterprise applications.
As a result, leaders are faced with capturing, ingesting, analyzing, storing, distributing,
securing the data, and transforming it into meaningful, valuable information.
Since 2000, the amount of information the federal government captures has increased
exponentially. In 2009, the U.S. Government produced 848 petabytes of data1 and U.S.
healthcare data alone reached 150 exabytes2 . Five exabytes (10^18 gigabytes) of data
would contain all words ever spoken by human beings on earth. At this rate, Big Data for
U.S. healthcare will soon reach zetabyte (10^21 gigabytes) scale and soon yottabytes
(10^24 gigabytes).
Yet, the mind-boggling volume of data that the federal government receives makes
information overload a fundamental challenge. In this expansion of data, there exists
new information that either has not been discoverable, or simply did not exist before. The
question is how to effectively capture new insight. Big Data properly managed, modeled,
shared, and transformed provides an opportunity to extract new insights, and make
decisions in a way simply not possible before now. Big Data provides the opportunity to
transform the business of government by providing greater insight at the point of impact
and ultimately better serving the citizenry, society and the world.

1 Source: IDC, US Bureau of Labor Statistics, McKinsey Global Institute Analysis
2 Roger Foster, “How to Harness Big Data for Improving Public Health,” Government Health IT, April 3, 2012, at http://www.govhealthit.com/news/how-harness-big-data-improving-public-health


9

Defining Big Data

Simply put, government leaders find
themselves stuck between a rock and a hard
place while facing ever-intensifying mission
and business challenges, the explosion in
the data available, and outmoded, out dated
information management capabilities that
simply limit their ability to respond. Some
of the questions government leaders face
include:

Big Data Definition
Although the term “Big Data” has become increasingly common,
its meaning is not always clear. For the purposes of this report,
the Commission tapped its collective experience, interviewed
government leaders from across the ecosystem, and arrived at
the definition below:

• How do I capture, manage and
exploit all this new data?
• How do I secure and govern it?
• How do I improve cross-organizational
information sharing for broader
connected intelligence?
• How do I build trust in the data,
through greater understanding of
provenance and lineage tied back to
validated, trusted sources?
• What advanced visualization
techniques, tools, and formats are
available for presenting information
to enable quick analysis and to create
new insights?
• Finally, how do I bridge the gap
in talent and human capital to take
advantage?

“Big Data is a term that describes
large volumes of high velocity,
complex and variable data that
require advanced techniques and
technologies to enable the capture,
storage, distribution, management,
and analysis of the information.”

Characteristics of Big Data
Big Data is often characterized by three factors: volume, velocity,
and variety. Fifteen percent of the information today is structured
information, or information that is easily stored in relational
databases of spreadsheets, with their ordinary columns and rows.
Unstructured information, such as email, video, blogs, call center
conversations, and social media, makes up about 85% of data
generated today and presents challenges in deriving meaning with
conventional business intelligence tools. Information-producing
devices, such as sensors, tablets, and mobile phones continue
to multiply. Social networking is also growing at an accelerated
pace as the world becomes more connected. Such information
sharing options represents a fundamental shift in the way people,
government and businesses interact with each other.
The characteristics of Big Data will shape the way government
organizations ingest, analyze, manage, store, and distribute
data across the enterprise and across the ecosystem. Table
1 illustrates characteristics of Big Data that more completely
describe the difference of “Big Data” from the historical
perspective of “normal” data.

10


Defining Big Data

Table 1: Characteristics of Big Data

Characteristic
Volume

Velocity

Description

Attribute

The sheer amount of data
generated or data intensity that
must be ingested, analyzed,
and managed to make
decisions based on complete
data analysis

According to IDC’s Digital
Universe Study, the world’s
“digital universe” is in the
process of generating 1.8
Zettabytes of information - with
continuing exponential growth
– projecting to 35 Zettabytes
in 2020

Increase in data sources,
higher resolution sensors

How fast data is being
produced and changed and the
speed with which data must
be received, understood, and
processed

• Accessibility: Information
when, where, and how the
user wants it, at the point of
impact

• Increase in data sources

• Applicable: Relevant,
valuable information for an
enterprise at a torrential
pace becomes a real-time
phenomenon

Driver

• Improved thru-put
connectivity
• Enhanced computing
power of data generating
devices

• Time value: real-time
analysis yields improved datadriven decisions

Variety

The rise of information coming
from new sources both inside
and outside the walls of the
enterprise or organization
creates integration,
management, governance, and
architectural pressures on IT

• Structured – 15% of data
today is structured, row,
columns
• Unstructured – 85% is
unstructured or human
generated information
• Semistructured – The
combination of structured and
unstructured data is becoming
paramount.

• Mobile
• Social Media
• Videos
• Chat
• Genomics
• Sensors

• Complexity – where data
sources are moving and
residing

Veracity

The quality and provenance of
received data

The quality of Big Data may
be good, bad, or undefined
due to data inconsistency
incompleteness, ambiguities,
latency, deception, model
approximations


Data-based decisions require
traceability and justification

11

Defining Big Data

Mission / Value of Big Data
The manner in which Big Data can be used to
create value across the government and in the
global economy is broad and far reaching. We are
at the cusp of a tremendous wave of innovation,
productivity, and growth – all driven by Big Data
as citizens, companies, and government exploit its
potential.
But, why should this be the case now? Hasn’t data
always been a part of the impact of information
technology? Yes, but the scale and scope of value
that data can bring is coming to an inflection point,
set to expand greatly as the availability of Big Data
converges with the ability to harness it and the
affordability of doing so. The federal government
has a significant opportunity to boost the efficiency
and value of investments for citizens at a time
when public finances are constrained and likely to
remain so, as U.S. and world population ages and
diversifies, the economy continues to globalize, and
the expectations on the part of the citizenry intensify.
Many key tenets for “Good Government” and
Big Data overlap. At its core, Big Data enables
government organizations to be smarter to improve
the productivity of the enterprise, and to serve the
needs of their stakeholders by improving decisionmaking in individual agencies and across the
government ecosystem. The ability to drive critical,
accurate insights with speed across a variety of data
sets and across a variety of channels (e.g., web
portals, internal systems, smart phones, tablets)
will dramatically improve a broad range of critical
government policies and practices.

Beyond the mission and business value derived directly by the
government’s capture, management, analysis, and storage
of Big Data, such efforts will also create new markets and
business opportunities for the private sector. Harnessing Big
Data will enable businesses to improve market intelligence,
thereby enhancing the value they provide to consumers. It will
also help to reduce business uncertainty for small businesses.
A recent report by McKinsey Global Institute (MGI) finds that
leveraging Big Data for insights can create significant value
for the economy, enhancing productivity and competitiveness
for companies and the public sector, and creating a
substantial economic surplus for consumers.
As Big Data becomes an increasingly valuable asset to the
government, the government’s embrace of the principles of
Big Data will lay the foundation for:
• Replacing or supporting human decision-making with
automated algorithms
• Reducing inefficiencies within an agency
• Creating transparency
• Improving performance by enabling experimentation to
discover needs and expose variability
• Improving ROI for IT investments
• Improved decision-making and operational intelligence
• Providing predictive capabilities to improve mission
outcomes
• Reducing security threats and crime
• Eliminating waste, fraud, and abuse
• Innovating new business models and stakeholder
services

12


Defining Big Data

The Use of Big Data: Potential Business Mission Value
Although the Big Data challenge is daunting, it is not insurmountable, and the opportunity is compelling.
There are many possibilities and approaches to managing and leveraging Big Data to address the mission
of government inclusive of stakeholder requirements. Ultimately, this report will provide guidance for a
framework to extract the Big Data needed to analyze and use for effective decision making. This will be
the baseline for a continuous feedback process to improve upon the outcomes identified or potentially
eliminate programs that are not delivering on the desired outcomes.
The potential applications of Big Data described below serve to illustrate the “Art of the Possible” in the
potential value that can be derived. These applications are consistent with the recently published Digital
Government Strategy.3 They require a customer focus and the ability to reuse and leverage data in
innovative ways.

Healthcare Quality and Efficiency
The ability to continuously improve quality and efficiency in the delivery of healthcare while reducing
costs remains an elusive goal for care providers and payers, but also represents a significant opportunity
to improve the lives of everyday Americans. As of 2010, national health expenditures represent 17.9%
of gross domestic product, up from 13.8% in 2000.4 Coupled with this rise in expenditures, certain
chronic diseases, such as diabetes, are increasing in prevalence and consuming a greater percentage of
healthcare resources. The management of these diseases and other health-related services profoundly
affects our nation’s well-being.
Big Data can help. The increased use of electronic health records (EHRs) coupled with new analytics
tools presents an opportunity to mine information for the most effective outcomes across large populations.
Using carefully de-identified information, researchers can look for statistically valid trends and provide
assessments based upon true quality of care.

Healthcare Early Detection
Big Data in health care may involve using sensors in the hospital or home to provide continuous monitoring
of key biochemical markers, performing real time analysis on the data as it streams from individual highrisk patients to a HIPAA-compliant analysis system. The analysis system can alert specific individuals and
their chosen health care provider if the analysis detects a health anomaly, requiring a visit to their provider
or a “911” event about to happen. This has the potential to extend and improve the quality of millions of
citizens’ lives.

3 http://www.whitehouse.gov/sites/default/files/omb/egov/digital-government/digital-government-strategy.pdf
4 http://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/NationalHealthExpendData/Downloads/tables.pdf


13

Defining Big Data

Transportation
Through improved information and autonomous features, Big Data has the potential to transform transportation in many ways.
The nemesis of many American drivers, traffic jams waste energy, contribute to global warming and cost individuals time and
money. Distributed sensors on handheld devices, on vehicles, and on roads can provide real-time traffic information that is
analyzed and shared. This information, coupled with more autonomous features in cars can allow drivers to operate more
safely and with less disruption to traffic flow. This new type of traffic ecosystem, with increasingly connected “intelligent cars,”
has the potential to transform how we use our roadways.5

Education
Big Data can have a profound impact on American education and our competitiveness in the global economy. For example,
through in-depth tracking and analysis of on-line student learning activities – with fine grained analysis down to the level of
mouse clicks – researchers can ascertain how students learn and the approaches that can be used to improve learning.
This analysis can be done across thousands of students rather than through small isolated studies.6 Courses and teaching
approaches, online and traditional, can be modified to reflect the information gleaned from the large scale analysis.

Fraud Detection – Healthcare Benefits Services
Big Data can transform improper payment detection and fundamentally change the risk and return perceptions of individuals
that currently submit improper, erroneous or fraudulent claims. For example, a significant challenge confronting the Centers
for Medicare and Medicaid Services (CMS) is managing improper payments under the Medicare Fee-For-Service Program
(FFS). The FFS distributes billions of dollars in estimated improper payments.7 Currently, contractors and employees identify
improper payments by selecting a small sample of claims, requesting medical documentation from the provider who submitted
the claims, and manually reviewing the claims against the medical documentation to verify the providers’ compliance with
Medicare’s policies.
This challenge is an opportunity to explore a use case for applying Big Data technologies and techniques, to perform
unstructured data analytics on medical documents to improve efficiency in mitigating improper payments. Automating
the improper payment process and utilizing Big Data tools, techniques and governance processes would result in greater
improper payment prevention or recovery. Data management and distribution could be achieved through an image
classification workflow solution to classify and route documents. Data analytics and data intelligence would be based
on unstructured document analysis techniques and pattern matching expertise.
The benefit is that the culture of submitting improper payments will be changed. Big Data tools, techniques and governance
processes would increase the prevention and recovery dollar value by evaluating the entire data set and dramatically
increasing the speed of identification and detection of compliance patterns.

5 http://www.forbes.com/sites/toddwoody/2012/09/19/automakers-on-the-road-to-self-driving-cars/
6 http://www.nytimes.com/2012/03/29/technology/new-us-research-will-aim-at-flood-of-digital-data.html
7 http://www.mcknights.com/gao-medicare-spent-48-billion-on-improper-fee-for-service-payments/article/208592/

14


Defining Big Data

Cyber Security

Weather

Government agencies face numerous challenges
associated with protecting themselves against cyber
attacks, such as managing the exponential growth in
network-produced datat, database performance issues
due to lack of ability to scale to capture this data, and
the complexity in developing and applying analytics
for fraud to cyber data. Agencies continue to look at
delivering innovative cyber analytics and data intensive
computing solutions. Cyber intelligence and other
machine generated data are growing beyond the limits of
traditional database and appliance vendors. Therefore,
requirements exist for fast data ingestion, data sharing,
and collaboration.

The ability to better understand changes in the frequency,
intensity, and location of weather and climate can benefit
millions of citizens and thousands of businesses that rely
upon weather, including farmers, tourism, transportation,
and insurance companies. Weather and climate-related
natural disasters result in tens of billions of dollars in losses
every year and affect the lives of millions of citizens. Much
progress has been made in understanding and predicting
weather, but it’s far from perfect. New sensors and analysis
techniques hold the promise of developing better long term
climate models and nearer term weather forecasts.

Agencies are looking to incorporate multiple streams
of data to benefit both human and automated analysis.
Cyber data such as host, network, and information
from the World Wide Web, are being fused with human
oriented information such as psychosocial, political,
and cultural information to form a more complete
understanding of our adversaries, motives, and
social networks. In addition, deep forensics, critical
infrastructure protections, Supervisory Control and
Data Acquisition (SCADA) security, and insider threat
protection are areas of focus for Big Data cyber security
capabilities.

New Ways of Combining
Information – Helping the
Unemployed Find Work

Fraud Detection –
Tax Collection
By increasing the ability to quickly spot anomalies,
government collection agencies can lower the “tax
gap” – the difference between what taxpayers owe and
what they pay voluntarily – and profoundly change the
culture of those that would consider attempting improper
tax filings. Most agencies practice a “pay and chase”
model, in which they accept returns and often pay out
tax refunds, and only ex post facto review a sampling
of returns in order to reveal unintentional or intentional
underpayment. Big Data offers the ability to improve
fraud detection and uncover noncompliance at the time
tax returns are initially filed, reducing the issuance of
questionable refunds.

Consistent with the Government Digital Strategy of making
information available and having a customer focus, Big
Data provides an opportunity to develop thousands of
new innovative ways of developing solutions for citizens.
For example, Big Data may be useful in proactively
helping unemployed job seekers find new opportunities
by combining their job qualifications with an analysis and
mining of available job opportunities that are posted on the
Internet (e.g., company Websites, commercial postings).
It’ll take some innovation, piloting and experimentation
to make these new ideas work, but the payoff can be
significant for individuals and society.

Conclusions
Government agencies should think about Big Data not as
an IT solution to solve reporting and analytical information
challenges but rather as a strategic asset that can be used
to achieve better mission outcomes, and conceptualized in
the strategic planning, enterprise architecture, and human
capital of the agency. Through this lens, government
agencies should create an ownership structure for the
data, treating it like any other asset – one that is valued
and secured. Ultimately, agencies should strive to address
the following two questions – “How will the business of
government change to leverage Big Data?” and “How will
legacy business models and systems be disrupted?”


15


The Commission has compiled a set of 10 case studies detailing the business or mission
challenge faced, the initial Big Data use case, early steps the agency took to address the
challenge and support the use case, and the business results. Although the full text of these case
studies will be posted at the TechAmerica Foundation Website, some are summarized below.

Table 2 – Case Studies High Level Summary
Agency/Organization/
Company
Big Data Project Name

Underpinning
Technologies

Big Data
Metrics

Initial Big Data
Entry Point

Public/User Benefits

Case Studies and Use Cases

National Archive and Records
Administration (NARA)
Electronics Records Archive

Metadata, Submission,
Access, Repository, Search
and Taxonomy applications
for storage and archival
systems

Petabytes,
Terabytes/sec,
Semi-structured

Warehouse
Optimization,
Distributed Info Mgt

Provides Electronic Records Archive
and Online Public Access systems for
US records and documentary heritage

TerraEchos
Perimeter Intrusion Detection

Streams analytic software,
predictive analytics

Terabytes/sec

Streaming and Data
Analytics

Helps organizations protect and monitor
critical infrastructure and secure
borders

Royal Institute of Technology of
Sweden (KTH)
Traffic Pattern Analysis

predictive analytics

Gigabits/sec

Streaming and Data
Analytics

Improve traffic in metropolitan areas by
decreasing congestion and reducing
traffic accident injury rates

Vestas Wind Energy
Wind Turbine Placement
Maintenance

Apache Hadoop

Petabytes

Streaming and Data
Analytics

Pinpointing the optimal location for wind
turbines to maximize power generation
and reduce energy cost

University of Ontario (UOIT)
Medical Monitoring

predictive analytics,
supporting Relational
Database

Petabytes

Streaming and Data
Analytics

Detecting infections in premature
infants up to 24 hours before they
exhibit symptoms

National Aeronautics and Space
Administration (NASA)
Human Space Flight Imagery

Metadata, Archival, Search
and Taxonomy applications
for tape library systems,
GOTS

Petabytes,
Terabytes/sec,
Semi-structured

Warehouse
Optimization

Provide industry and the public with
some of the most iconic and historic
human spaceflight imagery for scientific
discovery, education and entertainment

AM Biotechnologies (AM
Biotech)
DNA Sequence Analysys for
Creating Aptamers

Cloud-based HPC genomic
applications and
transportable data files

Gigabytes, 10
DNA sequences
compared

Streaming Data
Analytics, Warehouse
Optimization,

Creation of a unique aptamer
compounds to develop improved
therapeutics for many medical
conditions and diseases

National Oceanic and
Atmospheric Administration
(NOAA)
National Weather Service

HPC modeling, data from
satellites, ships, aircraft and
deployed sensors

Petabytes,
Terabytes/sec,
Semi-structured,
ExaFLOPS,
PetaFLOPS

Streaming Data
Optimization,

Provide weather, water, and climate
data, forecasts and warnings for the
protection of life and property and
enhancement of the national economy

Internal Revenue Service (IRS)
Compliance Data Warehouse

Columnar database
architecture, multiple
analytics applications,
descriptive, exploratory, and
predictive analysis

Petabytes

Streaming Data
Optimization,

Provide America's taxpayers top quality
service by helping them to understand
and meet their tax responsibilities and
enforce the law with integrity and
fairness to all

Centers for Medicare
Medicaid Services (CMS)
Medical Records Analytics

Columnar and NoSQL
databases, Hadoop being
looked at, EHR on the front
end, with legacy structured
database systems (including
DB2 and COBOL)

Petabytes,
Terabytes/day

Streaming Data
Optimization,

Protect the health of all Americans and
ensure compliant processing of
insurance claims

16

7



The case studies represent systems that have been in production. Given their maturity, some
case studies identify existing and new challenges that will require even greater evolution
of their IT infrastructures and technology to handle the level of compliance, retention and
accuracy required by current and new policies and by public and economic needs.
These new challenges create new use cases for needed Big Data solutions. These new use
cases demonstrate how Big Data technologies and analytics are enabling new innovations
in scientific, health, environmental, financial, energy, business and operational sectors. That
said, the men and women involved with the systems described in these case studies and use
cases are persevering. They are committed to the positive outcomes that Big Data solutions
offer: better protection, increased collaboration, deeper insights, and new prosperity. The
use cases will demonstrate that using the current and emerging technologies for Big Data
(including cloud-enabled Big Data applications) will drive new solutions to deliver insights
and information that benefits both the government and the public, thus enabling real-time
decision making.

National Archive and Records Administration (NARA)
The National Archive and Records Administration (NARA) has been charged with providing the Electronic Records Archive
(ERA) and Online Public Access systems for U.S. records and documentary heritage. As of January 2012, NARA is managing
manage about 142 terabytes (TB) of information (124 TB of which is managed by ERA), representing over 7 billion objects,
incorporating records from across the federal agency ecosystem, Congress and several presidential libraries, reaching back
to the George W. Bush administration. It sustains over 350 million annual online visits for information. These numbers are
expected to dramatically increase as agencies are mandated to use NARA in FY2012.
In addition to ERA, NARA is currently struggling to digitize over 4 million cubic feet of traditional archival holdings, including
about 400 million pages of classified information scheduled for declassification, pending review with the intelligence
community. Of that backlog, 62% of the physical records stored run the risk of never being preserved.
The NARA challenge represents the very essence of Big Data – how does the agency digitize this huge volume of
unstructured data, provide straightforward and rapid access, and still effectively governing the data while managing access
in both classified and declassified environments.
NARA has adopted an approach that put it on the path to developing the Big Data capability required to address its challenge.
This approach combines traditional data capture, digitizing, and storage capabilities with advanced Big Data capabilities
for search, retrieval, and presentation, all while supporting strict security guidelines. Dyug Le, Director of ERA Systems
Engineering writes, “It is best that the Electronic Records Archive be built in such a way so as to fit in a technology ecosystem
that can evolve naturally, and can be driven by the end users in ways that naturally ride the technology waves.” 8 The result
is faster ingestion and categorization of documents, an improved end user experience and dramatically reduced storage
costs. NARA notes the importance of record keeping in their drive toward electronic adoption and the cost benefit. It states,
“Electronic records can be duplicated and protected at less cost than paper records.” 9

8 http://ddp.nist.gov/workshop/ppts/01_05_Dyung_Le%20US_DPIF_NIST%20Digital%20Preservation%20Workshop.pdf
9 http://www.archives.gov/records-mgmt/policy/prod1afn.html


17


Royal Institute of Technology of Sweden (KTH)
Researchers at KTH, Sweden’s leading technical university, wanted to gather in real-time a wide array of data that might
affect traffic patterns, in order to better managed congestion. This real-time sensor data includes GPS from large numbers of
vehicles, radar sensors on motorways, congestion charging, weather and visibility etc. The challenge was collecting the wide
variety of data at high velocity and assimilating it in real time for analysis.
Collected data is now flowing into a commercial off-the-shelf (COTS) Streams Analytics software, a unique software tool
that analyzes large volumes of streaming, real-time data, both structured and unstructured. The data is then used to help
intelligently identify current conditions, and estimate how long it would take to travel from point to point in the city, offer advice
on various travel alternatives, such as routes, and eventually help improve traffic in a metropolitan area.
The KTH Big Data Deployment:
• Uses diverse data, including GPS locations, weather conditions, speeds and flows
from sensors on motorways, incidents and roadworks
• Enters data into the Streams Analytics software, which can handle all types of data,
both structured and unstructured
• Handles, in real time, the large traffic and traffic-related data streams to enable
researchers to quickly analyze current traffic conditions and develop historical
databases for monitoring and more efficient management of the system
The result has been a decrease in traffic congestion and accidents in the target cities. KTH is now looking to expand the
capability to support routing of emergency services vehicles.

18



Vestas Wind Energy
Since 1979, this Danish company has been engaged in the development, manufacture, sale, and maintenance of wind power
systems to generate electricity. Today, Vestas installs an average of one wind turbine every three hours, 24 hours a day,
and its turbines generate more than 90 million megawatt-hours of energy per year, enough electricity to supply millions of
households.
Making wind a reliable source of energy depends greatly on the placement of the wind turbines used to produce electricity in
order to optimize the production of power against wear and tear on the turbine itself.
For Vestas the process of establishing a location starts with its wind library, which combines data from global weather systems
with data collected from existing turbines. Data is collected from 35,000 meteorological stations scattered around the world
and from Vestas’s turbines. The data provides a picture of the global flow scenario, which in turn leads to mesoscale models
that are used to establish a huge wind library that can pinpoint the weather at a specific location at a specific time of day.
The company’s previous wind library provided detailed information in a grid pattern with each grid measuring 27x27 kilometers
(about 17x17 miles). Using computational fluid dynamics models, Vestas engineers can then bring the resolution down even
further—to about 10x10 meters (32x32 feet)—to establish the exact wind flow pattern at a particular location. However, in any
modeling scenario, the more data and the smaller the grid area, the greater the accuracy of the models. As a result, Vestas
wanted to expand its wind library more than 10 fold to include a larger range of weather data over a longer period of time.
To succeed, Vestas uses one of the largest supercomputers worldwide, along with a new Big Data modeling solution, to
slice weeks from data processing times and support 10 times the amount of data for more accurate turbine placement
decisions. Improved precision provides Vestas customers with greater business case certainty, quicker results, and
increased predictability and reliability in wind power generation.
• Reduces response time for wind forecasting information
by approximately 97 percent (from weeks to hours) to help
cut development time
• Improves accuracy of turbine placement with capabilities
for analyzing a greater breadth and depth of data
• Lowers the cost to customers per kilowatt hour produced
and increases customers’ return on investment
• Reduces IT footprint and costs, and decreases energy
consumption by 40 percent


19


University of Ontario Institute of Technology
The rapid advance of medical monitoring technology has done wonders to improve patient outcomes. Today, patients
routinely are connected to equipment that continuously monitors vital signs, such as blood pressure, heart rate and
temperature. The equipment issues an alert when any vital sign goes out of the normal range, prompting hospital staff to take
action immediately.
Many life-threatening conditions do not reach critical level right away, however. Often, signs that something is wrong begin
to appear long before the situation becomes serious, and even a skilled and experienced nurse or physician might not be
able to spot and interpret these trends in time to avoid serious complications. One example of such a hard-to-detect problem
is nosocomial infection, which is contracted at the hospital and is life threatening to fragile patients such as premature
infants. According to physicians at the University of Virginia, an examination of retrospective data reveals that, starting
12 to 24 hours before any overt sign of trouble, almost undetectable changes begin to appear in the vital signs of infants
who have contracted this infection. Although the information needed to detect the infection is present, the indication is very
subtle; rather than being a single warning sign, it is a trend over time that can be difficult to spot, especially in the fast-paced
environment of an intensive care unit.
The University of Ontario’s Institute of Technology partnered with researchers from a prominent technology firm that was
extending a new stream-computing platform to support healthcare analytics. The result was Project Artemis -- a highly flexible
platform that aims to help physicians make better, faster decisions regarding patient care for a wide range of conditions.
The earliest iteration of the project is focused on early detection of nosocomial infection by watching for reduced heart rate
variability along with other indications.
Project Artemis is based on Streams analytic software. An underlying relational database provides the data management
required to support future retrospective analyses of the collected data.
The result is an early warning that gives caregivers the ability to proactively deal with potential complications—such as
detecting infections in premature infants up to 24 hours before they exhibit symptoms. This system:
•
Holds the potential to give clinicians an unprecedented ability to
interpret vast amounts of heterogeneous data in real time, enabling
them to spot subtle trends
•
Combines physician and nurse knowledge and experience with
technology capabilities to yield more robust results than can be
provided by monitoring devices alone
•
Provides a flexible platform that can adapt to a wide variety of
medical monitoring needs

20



NASA Johnson Space Center
As the nucleus of the nation’s astronaut corps and home to International Space Station (ISS) mission operations, NASA
Johnson Space Center (JSC) plays a pivotal role in surpassing the physical boundaries of Earth and enhancing technological
and scientific knowledge to benefit all of humankind. NASA JSC manages one of the largest imagery archives in the world
and has provided industry and the public with some of the most iconic and historic human spaceflight imagery for scientific
discovery, education and entertainment. If you have seen it at the movies or on TV, JSC touched it first.
NASA’s imagery collection of still photography and video spans more for than half a century: from the early Gemini and Apollo
missions to the Space Station. This imagery collection currently consists of over 4 million still images, 9.5 million feet of
16mm motion picture film, over 85,000 video tapes and files representing 81,616 hours of video in analog and digital formats.
Eight buildings at JSC house these enormous collections and the imagery systems that collect, process, analyze, transcode,
distribute and archive these historical artifacts. NASA’s imagery collection is growing exponentially, and its sheer volume of
unstructured information is the essence of Big Data.
NASA’s human spaceflight imagery benefits the public through the numerous organizations that create media content for
social and public consumption. It is also used by the scientific and engineering community to avoid costly redesigns and to
conduct scientific and engineering analysis of tests and mission activities conducted at NASA JSC and White Sands Test
Facility.
NASA has developed best practices through technologies and processes to:
• Comply with NASA records retention schedules
and archiving guidance
• Migrate imagery to an appropriate storage medium
and format destined for the National Archives
• Develop technology to digital store down-linked images
and video directly to tape libraries
Key lessons learned have revolved around data lifecycle management and NASA’s Imagery Online records management tool
which contributes to the state of the art in records management.


21


Introduction
Powerful advances in new information management and business analytics technologies, like map reduce frameworks (such
as Hadoop, and Hadoop-like technologies), stream analytics, and massive parallel processing (MPP) data warehouses, have
been proven in deployment to support government agencies in harnessing the value from the increased volume, variety and
velocity that characterize Big Data. This section describes the different capabilities provided by these technologies, and how
they are combined in different ways to provide unique solutions. No single technology is required for a “Big Data Solution” –
they are not “must haves” – as initial Big Data Deployments are unique to the individual agency’s business imperatives and
use cases. The technologies can be placed, however, into the context of a broader enterprise Big Data model. The model
below highlights the ecosystem of technologies that can be used to support Big Data solutions, coupling legacy investments to
new technologies. As a result, the technologies listed in the model are not all new, but can be used as part of the solution set
(see Figure 1).

Figure1 Big Data Enterprise Model

Accelerators
Text Statistics
Internet / Intranet

Financial

Image/Video

Mining

Connectors

Geospatial
Times Series

Acoustic
Mathematical

Analytic Applications

Satellite

Realtime Analytics
Streaming

Map Reduce
Frameworks

Data Warehouse/
Database

Video/Pictures

Infrastructure
Mobile Devices
Provisioning

Configuration
Manager

Security

Cloud/
Virtualization

Storage
Infrastructure

Data
Compression

Job Tracking

Activity
Monitor

22

Job
Scheduling
Admin Tools

Documents

Workflow

Identity
Access Mgmt


Legacy IT Ecosystem

Social Sites

Integration Layer

Core Technologies


Big Data Infrastructure
A suitable technology infrastructure is a key prerequisite for embarking on a successful Big Data strategy. The proper
infrastructure is required to leverage the data that originates from the varied applications and data sources. Many government
agencies operate a diverse collection of systems based on a wide variety of technology architectures. Adjustments in data
center infrastructure may be necessary to implement a Big Data platform. For example, additional dedicated data storage
may be necessary to manage massive amounts of unstructured information. Network bandwidth is also a concern of current
environments.

Big Data Core Technologies
Real-Time Analytics and Streaming
performs the analysis of massive volumes of streaming
data, frequently with millisecond response times. The
data can originate from many sources: sensors, social
media sites, digital pictures, video, transaction records,
and communications surveillance and has a wide
variation in form. It is particularly ideal when users
need real-time decision support for high velocities of
data. The characteristics of real-time analytics and
streaming capabilities include rule-based algorithms,
large I/O capability, and parallel processing. It does not
necessarily require a significant amount of storage for
data at rest.

Map Reduce Frameworks are often based
upon Hadoop and Hadoop-like technologies. They
work by providing parallel processing capabilities that
move subsets of the data to distributed servers. The
primary use is processing massive amounts of data
in a scalable manner. Use cases include business
analytics, extraction, transformation and loading (ETL),
log analysis, and Web search engines. Most Hadoop
adopters rely on the native Hadoop Distributed File
System (HDFS) configured with direct attached storage
(DAS). HDFS is a distributed file system that provides
fault tolerant storage capable of managing distributed
data across many nodes. The systems commonly work
in a “map and reduce” paradigm that allows for a master
server to distribute (i.e., map) the problem to worker
servers, collect each server’s results, and reduce it to
a collective result. These technologies are “scale out”
and can accommodate large, unbounded data sets
without the format requirements of a relational database
management system (RDBMS).

Frequently, the software includes redundancy in
the file system (e.g., Hadoop File System) or error
checking in the data that can facilitate the use of
commodity hardware. As a result, it is well-suited
to deal with semi-structured and unstructured
data, and is particularly ideal when there is a large
amount of data maintained to discover new insights
over time and when there’s not a firm understanding
of the value or use of the data a priori. It is good
for finding a needle in the haystack among data
that may or may not be “core data” for an agency
today. Hadoop and Hadoop-like technologies tend
to work with a batch paradigm that is good for
many workloads, but is frequently not sufficient for
streaming analysis or fast interactive queries.

Data Warehousing can be part of an overall
integrated Big Data solution because it is ideal for
analyzing structured data from various systems
to deliver deep operational insight with advanced
in-database analytics. Traditional on-line analytic
processing (OLAP) data warehouses have scaleup characteristics, require relational, structured
data, and have a ceiling on the size of the data that
can be processed (e.g., when joining the data with
structured query language), but can be a powerful
source of analytical capability that can be integrated
to provide significant customer value. Frequently
the data in an OLAP-based data warehouse has
been pre-processed and is known to be of high
quality.


23


Big Data Accelerators
Accelerators are software applications,
connectors, or interfaces that provide
value-add capabilities such as
implementing analytics or enhancing
the ability to integrate. Accelerators can
decrease deployment time and speed the
time to realizing value from the Big Data
investment. Accelerators include:
• Text extraction tools or interfaces to
common Text Extraction or Natural
Language Processing products
• Financial tools and interfaces to
common financial products
• Geospatial support and interfaces
for ground station and ingest
solutions
• Geospatial integration and
dissemination
• Acoustic interfacing support
• Imagery and video mining, marking,
monitoring, and alerting capability or
interfacing support

Integration Layer
Integration between new Big Data capabilities and legacy
IT investments is often necessary for deploying Big Data
solutions. Information integration and governance allows
agencies to understand, cleanse, transform, manage, and
deliver trusted information to the critical business initiatives.
We recommend that agencies develop Big Data governance
plans. The plans should be a holistic approach to help guide
the agency from an information management perspective and
follow these key items:
• Identify data and content that are vital to its mission
• Identify how, when, where, and to whom information
should be made available
• Determine appropriate data management, governance,
and security practices
• Identify and prioritize the information projects that
deliver the most value
A Big Data information plan ensures that the major
components of the governance plan are working in unison.
There are four components to an effective information plan:
• Information strategy is the vision that guides decisions
and helps the organization determine how best to support
business goals
• Information infrastructure is the technology and
capabilities that are needed to establish a common
information framework
• Information governance is the policies and practices
that facilitate the management, usage, improvement, and
protection of information across the organization
• Road map is a phased execution plan for transforming
the organization

24



In practice, Big Data technologies can be integrated to provide a comprehensive solution for government IT leadership. Big
Data solutions are often used to feed traditional data warehousing and business intelligence systems. For example, Hadoop
within a Big Data model can be used as a repository for structured, semi-structured (e.g., log file data), and unstructured
data (e.g., emails, documents) that feeds an OLAP data warehouse. The analytics and visualization tools pull data from the
OLAP data warehouse and render actionable information through reports. However, the fundamental step of cleansing the
data prior to loading the OLAP data warehouse is absolutely critical in developing “trusted information” to be used within the
analytics.
Technologies are evolving to provide government IT leadership choices of characteristics and costs. For example, NoSQL
and Hadoop technologies include the ability to scale horizontally without a pre-defined boundary. These technologies
may run on commodity hardware or can be optimized with high-end hardware technology tuned specifically to support Big
Data. Similarly, NoSQL and Hadoop have different characteristics and capabilities than traditional RDBMSs and analysis
tools. The ideal enterprise data warehouse has been envisaged as a centralized repository for 25 years, but the time has
come for a new type of warehouse to handle Big Data. MapReduce, Hadoop, in-memory databases and column stores
don’t make an enterprise data warehouse obsolete. The new enterprise data warehouse will leverage all of these software
technologies in the RDBMS or via managed services. This “logical data warehouse” requires realignment of practices and a
hybrid architecture of repositories and services. Software alone is insufficient — it will demand the rethinking of deployment
infrastructures as well.
Scalable analytics using software frameworks can be combined with storage designs that support massive growth for costeffectiveness and reliability. Storage platform support for Big Data can include multi-petabyte capacity supporting potentially
billions of objects, high-speed file sharing across heterogeneous operating systems, application performance awareness and
agile provisioning. Agencies must consider data protection and availability requirements for their Big Data. In many cases
data volumes and sizes may be too large to back up through conventional methods. Policies for data management will need
to be addressed as well; the nature of many Big Data use cases implies data sharing, data reuse and ongoing analysis.
Security, privacy, legal issues such as intellectual property management and liability, and retention of data for historical
purposes need to be addressed.

Networking Considerations
Many Big Data implementations require the movement
of information across the network and between disparate
organizations that operate separate networks. This is
especially true when working with near real-time data sources.
Current networks may need remediation as information is
pulled into storage repositories from multiple sources. New
network loads may necessitate a restructuring of agencies’
networks to separate traffic used in Big Data solutions from
that of normal business operations. The good news is
that most government agencies are undergoing significant
data center transformations to consolidate operations and
standardize on common infrastructures. The impact of
standardization, especially at the network and physical level,
will allow data center operators to achieve greater flexibility in
how they deploy modular data centric solutions. Agencies will
gain efficiencies if they ensure data center transformations
address these necessary Big Data infrastructure requirements
as part of their consolidation strategy.


25


Cloud Computing Considerations
Consistent with federal CIO policy, Big Data presents federal IT leadership with options for deployment infrastructure that can
include cloud computing for development, test, integration, pilots and production. Cloud computing approaches can allow for
faster deployment, more effective use of scarce IT resources, and the ability to innovate more quickly. Innovation is enabled
through the dynamic use of low cost virtual environments that can be instantiated on demand. This allows organizations to
succeed quickly or fail quickly and incorporate their lessons learned. As with any deployment option, cloud computing should
be evaluated against the ability to meet application and architecture requirements within the cost structure of the organization.

Figure 2: Notional Information Flow – The Information Supply Chain

Source Data
Applications

Data Preparation

Data Transformation
Metadata Repository

Business Intelligence
Decision Support

Analysts

Data Mining
Statistics

Streaming Data
Text Data

Optimization
Simulation

Multi-Dimensional

Industry
Domain Expert

Semantic
Analysis

Time Series

Fuzzy
Matching

Analytics
Solution End
User

New
Algorithms

Geo Spatial

Other
Analysts and
Users

Video Image

Network
Algorithms

Relational
Social Network

Security, Governance, Privacy, Risk Management
Data Acquisition

Filtering, Cleansing,
and Validation

Storage, Hadoop,
Warehousing

Core Analytics

Users

A notional flow of information can be a helpful approach in understanding how the technologies underpinning the Big
Data Enterprise Model can come together to meet government’s needs. This flow follows a simple Understand, Cleanse,
Transform and Exploit model. At the end of the day, the key is to map available data sources through the analytic
process to the target use cases.

26



Understand Source Data and Applications
The first step in any information integration and transformation initiative – Big Data or otherwise – is to identify and
understand the relevant sources of data, their degree of volume, variety and velocity, and their level of quality. This
understanding helps determine the degree of difficulty in accessing the data, the level of transformation required, and
the core Big Data technologies that will enable to agency to manage and exploit it.

Data Preparation – Cleansing Verification
Once an agency understands data sources in the context of the target use case, it can begin to define the method and
manner of the data preparation required to support the target use case. For example, unstructured data may require a simple
pass through for direct analysis or it may be filtered, cleaned and used for downstream processing. Structured information
– such as addresses, phone numbers, and names – may require standardization and verification. Specifics depend on
operational and business requirements.

Data Transformation
Once the data fueling a Big Data use case has been cleansed and verified, agencies may consider additional transformation
to extract insight and enhance its value. Data may be available for direct analysis (e.g., through Hadoop) or may need
additional processing before it can be fully leveraged for the intended use case. Unstructured data, for instance, may
be broken down and rendered in a structured format – an agency may want to perform entity extraction to associate
organizational or individual names with specific documents. Further, agencies may seek to aggregate data sources, or to
understand the non-obvious relationships between them. The goal is trusted information – information that is accurate,
complete, and insightful – such that every nugget of information has been extracted from the base data.

Business Intelligence/Decision Support
Once trusted information has been established, agencies can then use the broadest range of analytic tools and techniques
to exploit it. These tools and techniques range from the most basic business intelligence capabilities, to more sophisticated
predictive analytics, to anomaly detection, content analytics, sentiment analytics, imagery, aural analytics and biometrics.
Once the data is brought into the Big Data environment, the critical step is to process it to glean new insights. For example,
Hadoop can be used to analyze unstructured data residing on many distributed compute instances or business intelligence
tools can be used to analyze a structured data warehouse.

Analysts/Visualization
The final step in the information supply chain is to deliver the new insights created in the previous steps in a manner that most
effectively supports the business requirements and use cases. This implies one of two approaches, depending on the users
and the business requirement. Either the collection of new insights is provided through a visualization or collaboration tool
that enables the users to explore the information, ask questions and uncover new insights; or, the information is simply used
to fuel existing work process applications to improve existing processes. Either way, the user needs to be provisioned with
data that meets the Big Data needs.


27


The TechAmerica Foundation Big Data Commission’s report describes the business and mission challenges government
agencies face, and the unique role that Big Data can play in addressing these challenges. The report defines the
characteristics of Big Data – Volume, Variety, and Velocity – and the key underpinnings required for government leaders
to address Big Data challenges and capture the opportunities it offers. This report seeks to illuminate the possibilities
by examining potential government use cases, and case studies describing successful deployments. In addition, the
Commission has described proven technology and governance principles, methods for getting things done, and ways of
addressing privacy and security concerns.

Observations Lessons Learned
So what have we learned? In our discussion with leaders from across the government
ecosystem, and examining the case studies, five central themes have emerged -- each
of which coincidentally is well aligned with established engineering best practices:

1. Define business requirements: Successful Big Data initiatives commonly start
with a set of specific and well defined mission requirements, versus a plan to deploy a
universal and unproven technical platform to support perceived future requirements.
The approach is not “build it and they will come,” but “fit for purpose.”

2. Plan to augment and iterate: Successful initiatives favor augmenting current
IT investments rather than building entirely new enterprise scale systems. The new
integrated capabilities should be focused on initial requirements but be part of a larger
architectural vision that can include far wider use cases in subsequent phases of
deployment. This approach provides government leaders with lightweight, low trauma
solutions to their most immediate pain points within a strategic context that will move
them towards an ever greater ability to leverage Big Data.

3. Big Data entry point: Successful deployments are characterized by three
“patterns of deployment” underpinned by the selection of one Big Data “entry point”
that corresponds to one of the key characteristics of Big Data. Some initiatives do indeed
leverage a combination of these entry points, but experience shows these
are the exception.
• Velocity: Use cases requiring both a high degree of velocity in data processing
and real time decision making, tend to require Streams as an entry point.
• Volume: Government leaders struggling with the sheer volume in the data they
seek to manage, often select as an entry point a database or warehouse architecture
that can scale out without pre-defined bounds.
• Variety: Those use cases requiring an ability to explore, understand and analyze
a variety of data sources, across a mixture of structured, semi-structured and
unstructured formats, horizontally scaled for high performance while maintaining low
cost, imply Hadoop or Hadoop-like technologies as the entry point.

28



4. Identify gaps: Once an initial set of business requirements have been identified and defined, government IT leaders
assess their technical requirements and ensure consistency with their long term architecture. Leaders should identify gaps
against the Big Data reference architecture described previously, and then plan the investments to close the gaps.

5. Iterate: From these Phase I deployments clients typically then expand to adjacent use cases, building out a more robust
and unified Big Data platform. This platform begins to provide capabilities that cut across the expanding list of use cases,
and provide a set of common services to support an ongoing initiative. Beyond adding in adjacent core capabilities – e.g.,
Streams, Hadoop, Warehousing and Streams, these services often include:
• Information Integration and Governance
• Privacy Security
• Common Metadata Management
• Visualization and Discovery

Recommended Road Map for Getting Started
Based on these observations, the Big Data Commission recommends the following
five step cyclical approach to successfully take advantage of the Big Data opportunity.
These steps are iterative versus serial in nature, with a constant closed feedback loop
that informs ongoing efforts. Critical to success is to approach the Big Data initiative
from a simple definition of the business and operational imperatives that the government
organization seeks to address, and a set of specific business requirements and use cases
that each phase of deployment with support. At each phase, the government organization
should consistently review progress, the value of the investment, the key lessons learned,
and the potential impacts on governance, privacy, and security policies. In this way, the
organization can move tactically to address near term business challenges, but operate
within the strategic context of building a robust Big Data capability.

Figure 3: Road Map for Getting Started

Define

Assess

Plan

Execute

Review


29


Execute

Plan

Assess

Define

Table 3: Practical Road Map Summary
Define the Big Data
opportunity including the
key business and mission
challenges, the initial use
case or set of use cases,
and the value Big Data can
deliver

•  Identify key business challenges, and potential use cases to address
•  Identify areas of opportunity where access to Big Data can be used to better
serve the citizenry, the mission, or reduce costs
•  Ask – does Big Data hold a unique promise to satisfy the use case(s)
•  Identify the value of a Big Data investment against more traditional analytic
investments, or doing nothing
•  Create your overall vision, but chunk the work into tactical phases (time to value
within 12-18 month timeframe)
•  Don’t attempt to solve all Big Data problems in the initial project – seek to act
tactically, but in the strategic context of your key business imperatives

Assess the organization’s
currently available data
and technical capabilities,
against the data and
technical capabilities
required to satisfy the
defined set of business
requirements and use
cases

•  Assess the use case across velocity, variety and volume requirements, and
determine if they rise to the level of a Big Data initiative, versus a more traditional
approach
•  Assess the data and data sources required to satisfy the defined use case,
versus current availability
•  Assess the technical requirements to support accessing, governing, managing
and analyzing the data, against current capability
•  Leverage the reference architecture defined in the report above to identify key
gaps
•  Develop an ROI assessment for the current phase of deployment (ROI used
conceptually, as the ROI may be better services for customers/citizens and not
necessarily a financial ROI)

Select the most appropriate
deployment pattern and
entry point, design the “to
be” technical architecture,
and identify potential policy,
privacy and security
considerations

• 
• 
• 
• 
• 
• 
• 
• 
• 

The gov’t agency deploys
the current phase Big Data
project, maintaining the
flexibility to leverage its
investment to accommodate
subsequent business
requirements and use cases

Identify the “entry point” capability as described in the section above
Identify successful outcomes (success criteria)
Develop architectural roadmap in support of the selected use case or use cases
Identify any policy, privacy and security considerations
Plan iterative phases of deployment
Develop program management and acquisitions planning
Identify required skills, resources and staffing
Plan development, test and deployment platforms (e.g., Cloud, HW)
If appropriate, Pilot to mitigate business and technical risk

•  Deploy the current phase project plan
•  Build out the Big Data platform as the plan requires, with an eye toward flexibility
and expansion
•  Deploy technologies with both the flexibility and performance to scale to support
subsequent use cases and corresponding data volume, velocity and variety

Review

The gov’t agency continually
reviews progress, adjusts the
deployment plan as required,
and tests business process,
policy, governance, privacy
and security considerations

30

•  This is a continual process that cuts across the remainder of the roadmap steps
•  Throughout the assess and planning stages, continually review plans against set
governance, privacy, security policies
•  Assess big data objectives against current Federal, state and local policy
•  At each stage, assess ROI, and derive lessons learned
•  Review deployed architecture and technologies against the needs of the broader
organization – both to close any gaps, as well as to identify adjacent business areas
that might leverage the developing Big Data capability
•  Move toward Big Data Transformation in a continual iterative process


Public Policy

Public Policy
Accelerating Uptake of Big Data in the Federal Government
As the federal government moves to leverage Big Data, it must look closely at current
policies and determine whether they are sufficient to ensure it can maximize its promise.
Issues as varied as procurement processes for acquiring Big Data solutions, research and
development funding and strategies, and workforce development policies will have a dramatic
effect on the success or limitations of federal Big Data efforts. Furthermore, understanding
and addressing citizens’ expectations on privacy and security is critical for government
to implement Big Data solutions successfully. The government should evaluate Big Data
policy issues with a view toward removing the unnecessary obstacles to Big Data use, and
driving specific actions to accelerate the use and dissemination of the government’s data
assets. In this section of the report, we address specific strategic policy areas where careful
consideration by policy makers can greatly accelerate uptake of Big Data efforts.
The culture of information sharing and decision making needs to grow to include Big Data
analysis. Without providing decision-makers the policy; and the incentives to use Big
Data for insights and predictions; and the guidance on the use and sharing of information,
government will not realize the tremendous value Big Data can offer.
The best private companies today are relentlessly data-driven. They have led the way in
industry in measuring their performance carefully and in using hard quantitative information to
guide plans and decisions. Unfortunately, in the federal government, daily practice frequently
undermines official policies that encourage sharing of information both within and among
agencies and with citizens. Furthermore, decision-making by leaders in Congress and the
Administration often is accomplished without the benefit of key information and without
using the power of Big Data to model possible futures, make predictions, and fundamentally
connect the ever increasing myriad of dots and data available. As recognized in a recent
Government Accountability Office (GAO) report,10 Congress may miss opportunities to use
performance information produced by agencies. Such data could be leveraged to enact
targeted law and policy. So too, the Administration may miss the chance to leverage realtime data as it fulfills program missions. Both branches of government stand to benefit from
enhancing policies directed at the use of Big Data approaches as part of their daily routine.

10 Managing for Results: A Guide for Using the GPRA Modernization Act to Help Inform Congressional Decision Making (gao.gov/assets/600/591649.pdf).


31

Public Policy

The federal government should examine existing organizational and technical
structures to find and remove barriers to greater Big Data uptake and, where
needed, take action to accelerate its use. Specifically, the government should:
• Assess if annual reviews should incorporate performance metrics based on
whether the use of value-added data collaboration activities would improve
department and agency performance. Such metrics could promote data
sharing and uptake of Big Data in every-day activities. The Administration
has made progress in utilizing a host of data-driven reviews to enhance
performance. Building upon the success of the TechStat IT program review
process, many agencies have instituted data-driven review programs aimed at
overarching business and service delivery challenges. HUDStat, FEMAStat,
and similar efforts at the Social Security Administration serve as good
examples upon which to build. Agencies use these approaches primarily on a
quarterly basis. Their use should be expanded.
• Expand upon requirements in the Digital Government Strategy to share
and expose data assets of federal agencies to the public and private sector.
Requirements to share are beginning to take hold across federal agencies,
but stronger requirements and evaluation of agency compliance will help drive
further efforts ahead.
• Name a single official both across government and within each agency
to bring cohesive focus and discipline to leveraging the government’s data
assets to drive change, enhance performance, and increase competitiveness.
To be certain the value of data is properly harnessed across organizational
boundaries, Big Data should not be seen as the exclusive domain of the IT
organization or CIO.
• Examine innovative agency approaches to focus on Big Data
organizationally, such as the Federal Communications Commission’s (FCC)
decision to appoint a Chief Data Officer at the FCC, and whether such
approaches would be appropriate at other agencies.11
These efforts should take a strategic view of the importance of leveraging Big
Data for new ways of optimizing operations. It will also be important to clearly
define roles and responsibilities to ensure that the discipline of Big Data is not
overlooked, but rather is integrated into daily agency operations and decisionmaking.

11 www.fcc.gov/data/chief-data-officers

32


Public Policy

Education
and Workforce Development
The ability to glean new insights from the massive amounts
of data demands new skills and “out-of-the-box” thought
processes. At the same time, competition for these new
skills and key talent is increasing. Growing demand and
competition for top talent, particularly in science, technology,
engineering, math, and analytic fields has made hiring and
workforce retention increasingly difficult. Recent workforce
analyses indicate that there are looming talent shortfalls in
key IT disciplines, including those required to implement and
manage evolving technologies, such as those associated with
Big Data. We recommend a strategy to leverage new talent,
while increasing and retaining existing talent with Big Data
skills.

Leveraging New Talent
Our greatest source of new thinkers is the collection of
students coming from colleges and universities around the
country. For government to leverage this source of creative
thinkers, the internship programs that currently exist could
be expanded to include areas of specific data analytical
focus that demand new ways of addressing difficult problems
by instituting project focused internship assignments. The
additional advantage of such a program would be the
increasing connection of government internal issues with
people who, as they go out into private industry, can leverage
their learning and experience as interns to identify and bring
back to government new innovations and opportunities for
applying the data that abounds across government agencies.

Increasing Talent
The White House Office of Management and
Budget (OMB) should create a formal career track
for IT managers and establish an IT Leadership
Academy to provide Big Data and related IT
training and certification. This academy should
foster intellectual curiosity – a trait required to
shift workforce focus from transition processing to
analytic thinking and continual improvement. To
complement these OMB initiatives, agencies should
cross-train IT managers to understand functional
disciplines so that Big Data and related IT solutions
can be effectively utilized to support the mission.
We recommend that the IT Leadership Academy
be organized using established and proven
models of professional development, competencybased education, and in collaboration with select
academic centers. These centers can be selected
based on their interest in the Academy and their
accomplishments in professional development.
To help acquisition and IT personnel understand,
adopt, and use Big Data solutions, government
agencies, companies, and academia should
develop and disseminate educational resources and
programs aimed at educating the federal workforce
on the technical, business, and policy issues
associated with the acquisition, deployment, and
operation of Big Data solutions.
We also recommend a broader coalition between
Big Data Commission members, academic centers,
and professional societies to articulate and maintain
professional and competency standards for the field
of Big Data. Such standards will guide colleges and
universities across the country to develop relevant
programs, thus increasing the pool of qualified
students for Big Data related roles.


33

Public Policy

Research and Development Considerations
The Administration’s announcement of $200 million in
Big Data RD funding demonstrates its commitment
to the transformative nature of Big Data solutions.
These investments are a useful start, but much more is
needed. In much the same way that sustained federal
investment in the technologies underlying the Internet
spawned an entirely new technology and economic
ecosystem, continued and aggressive investment in the
Big Data discipline is critical to the development of new
tools, economic models, and educational approaches to
advancing Big Data uptake and utilization.
The Commission recommends that the OSTP further
develop a national research and development strategy
for Big Data that encourages research into new
techniques and tools, and that explores the application of
those tools to important problems across varied research
domains. Such research could be facilitated, and cost
could be optimized, by considering the establishment of
experimental laboratories within which agencies could
explore and test new Big Data technologies without the
need to spend funds on their own. This strategy should
not be limited to the major research organizations in
the federal government, but rather focus on the roles
and skill sets of all levels of government and create
incentives for the private sector to innovate and develop
transformative solutions.

The RD strategy should also move beyond technical
attributes and solutions, to address the wide range of
domains where improvements are needed. Among the
issues we recommend including in the national Big Data
RD strategy are the following:
A. Education and training advancements, including
novel approaches to curriculum development and
data-intensive degree programs and scholarships to
expand the preparation of new generations of data
scientists;
B. New management and analytic tools to address
increasing data velocity and complexity;
C. Novel privacy-enhancing and data management
technologies;
D. Development of new economic models to
encourage data sharing and monetization of Big Data
solutions and collaborative data science efforts;
E. Continued research and development of advanced
computing technologies that can effectively process,
not only the vast amounts of data being continually
generated, but also the various types of data; and
F. Policy guidelines that mandate that agencies
identify clear objectives for use of Big Data
technologies with a focus on metrics that can
reduce the cost of operations.
The Commission has spent time with the crossgovernment Big Data Steering Group operating under
the auspices of the National Coordination Office for
Networking and Information Technology Research
and Development (NITRD). We applaud their efforts
to explore additional avenues to innovation, including
competitions and prizes, and encourage the RD
strategy to address this approach explicitly. The
Administration has achieved success with internal
government competitions, and we encourage efforts
of a similar nature.

34


Public Policy

Privacy Issues
As the earlier best practice section reveals, good data
protection practices are in place, as evidenced by the
existing policies and industry best practices related to
privacy and security by design context, accountability,
and transparency. The Commission does not believe
that realizing the promise of Big Data requires the
sacrifice of personal privacy. Accordingly, education
and execution on existing privacy protections and
procedures must be part of any Big Data project.

1. Ways to simplify the understanding of privacy management
obligations through the issuance of one guiding set of
principles and practices to be applied by all agencies. Along
with the streamlining of access to what management practices
are required, this guidance also should explicitly note the
penalties for violating privacy requirements.

There are over 40 laws12 that provide various forms of
privacy protections. Some, such as the Privacy Act
of 1974, provide protections to personal data used by
government; others are more sector-specific, such as
the Health Insurance Portability and Accountability Act
(HIPAA), and the Financial Modernization Act (GrammLeach-Bliley Act 1999), which concern health and
financial information, respectively. Further guidance
from OMB, and greater collaboration with industry and
stakeholders, on applying these protections to current
technology and cultural realities could help accelerate
the uptake of Big Data. Specifically, we recommend
that OMB strongly collaborate with industry and
advocacy groups and issue additional guidance that
addresses:

3. The growing concern about the use, aggregation, sharing,
retention, and deletion of data with an eye toward identifying
best practices that may be needed to realize Fair Information
Practice Principles (FIPPs).

2. The government’s use of data that is not part of a Privacy
Act System of Records.

4. The need to recognize and promote citizen confidence by
communicating clearly that, for many of Big Data projects,
the data being aggregated are non-regulated, de-identified
information with no need to re-identify to derive benefits.
Of particular note should be the fact that data sets can be
used without personally identifiable information (PII) and still
yield positive results (e.g., disease spread predictions).
5. The use of clear public policies regarding notice and
access for Big Data to enable choice and transparency, as
articulated by the FIPPs. Along these lines, stakeholders
could explore a “layered notice” approach, where basic
information is available for all, with notice/consent/access
requirements scaled to instances where more information is
sought.
Most importantly, leadership at all levels of government is
needed on privacy as over time with relatively little guidance,
agencies have taken action, or not, as they best saw fit.
Privacy guidance from OMB should be updated regularly to
take into account the latest technologies, cultural attitudes
and norms, and government needs. Fundamentally, any
policy must be drafted carefully to enable both innovation and
appropriate data protection.

12 http://healthit.hhs.gov/portal/server.pt/gateway/PTARGS_0_11113_911059_0_0_18/Federal%20Privacy%20Laws%20Table%202%2026%2010%20Final.pdf


35

Transforming Government with Big Data

Transforming Government with Big Data

Recommended

Recommended

More Related Content

What's hot

What's hot (17)

Similar to Transforming Government with Big Data

Similar to Transforming Government with Big Data (20)

More from Ajay Alex

More from Ajay Alex (10)

Recently uploaded

Recently uploaded (20)

Transforming Government with Big Data